### Use of UAVs to calculate with precision areas of cashew farms?

I am M&E for an intervention which would like to capture the number of hectares under specific techniques/technologies (Conservation / technical climate smart agricultural, plantation management Improved: better weeding and pruning technologies; Improved Utilization of varieties for grafting).... Which steps to follow to use UAVs to calculte with precision aeras of cashew farms?

The intervention areas consist of a large number of very small farms (more than 2,500), ranging from less than one hectare to over 20 hectares, with a total area of ​​about 5,000 hectares. These farms are located in 5 communes in a circumscribed circle with a radius of about 250 Km. Two months to evaluate the initial whole situation and also an assess of state of affairs should  occurs twice a year...

Under theses conditions, how many UAVs to perform these operations?

#### Replies

• The easy way is to get a multicopter with the sensing equipment you need, a car, an RTK and a software that assembles the data you gathered. Nothing new, nothing fancy, and you will be spending tons of money and time on getting to the operation site.

The other end of the option range is a UAS that flies 250km does 20Ha mapping and flies 250km back. While this is the cheaper to operate it is more expensive to buy/develope. ZTE is working on one now.

There is a 3rd solution which in turn will probably be the final winner on the long run, but I am not sure how far are they from deployment stage. https://www.youtube.com/watch?v=9yEl0-bCA9M.

• Wauw.. A radius of 250km.. That's almost 200000 km2 which is the size of a small country. We do around 2-4 km2 per flight with our fixed wings. Even if you can increase to lets say 20km2 if allowed to fly above 500m, just do the math you'll see uav will take ages.. I would go for a combination of quick scanning with satellite imagery to deliniate the interesting areas and then plane/uav depending on the size and operational restrictions. NIR camera will definitly help to improve the identification but for most you need smart algorithms based on textural extraction and paralel computing to analyse the imagery. I work for a company that can provide all that... Just pm me if interested.

• This requires very specific NIR sensors, as well as unique data processing and analysis. Most important, need minimum resolution, overlap, and must know the precise timing during the growing season. Our company specializes in creating these case studies. Private Message me at phil@agpixel.com if interested.